I want to know if my Pulse settings are correct. Correct point in not
using 20KHz, but something more in the listening range. Like 100Hz to
15KHz

Carlos


On Fri, Mar 28, 2025 at 1:41?PM Andy I via groups.io
<AI.egrps+io@...> wrote:

On Fri, Mar 28, 2025 at 12:32 PM, Carlos E. Martínez wrote:

This should be the first time I will be using a SW to test a RIAA preamp response.

I was thinking of copying the settings I use for testing power amps, but I am not sure it's correct.

They are: PULSE(-.4 .4 0 10n 10n 25u 50u 10)

Would they be fine for this new test?

Un, what are you asking??

What is an "RIAA square wave"?

What is your PULSE waveform intended for? That should lead you in the direction of answering your own question.

We can not read your mind. I have no idea whether your PULSE waveform is "fine" or not.

However, here is one little point about it: Your "square wave" is not a square wave. I am not referring to the fact that the rising and falling edges aren't instantaneous, which is of course impossible. I refer to the fact that it does not have a 50% duty cycle. The pulse is high for 25.01 microseconds, and low for 24.99 microseconds. To correct that, change the "Ton" time to 24.99 us, which is 25 us minus your 10 ns rise and fall time. Then it will be symmetrical, 50% high and 50% low.

Also, do you really need to limit the PULSE train to only 10 cycles? If you don't want to limit the number of cycles, leave that field (Ncycles) blank.

Andy

开云体育

This should be the first time I will be using a SW to test a RIAA preamp response.

?

I was thinking of copying the settings I use for testing power amps, but I am not sure it's correct.

?

They are: PULSE(-.4 .4 0 10n 10n 25u 50u 10)

?

Would they be fine for this new test?

Not sure what you hope to learn from 20kHz square wave testing of an RIAA preamp, but it would depend on whether the 75μs pole is implemented passively, or in the main feedback loop. Are you testing it with an inverse RIAA network?

Not that it probably matters, but your pulse specification isn't quite square. To correct that, it should be:

PULSE(-400m 400m 0 10n 10n {25u-10n} 50u 10)

--
Regards,
Tony

开云体育

It doesn't seem right to use a pulse signal for an audio application.

Hi,

?

This should be the first time I will be using a SW to test a RIAA preamp response.

?

I was thinking of copying the settings I use for testing power amps, but I am not sure it's correct.

?

They are: PULSE(-.4 .4 0 10n 10n 25u 50u 10)

?

Would they be fine for this new test?

?

Thanks!

?

Carlos

On 3/28/25 11:32 AM, Carlos E. Martínez via groups.io wrote:

Hi,
This should be the first time I will be using a SW to test a RIAA preamp response.
I was thinking of copying the settings I use for testing power amps, but I am not sure it's correct.
They are: PULSE(-.4 .4 0 10n 10n 25u 50u 10)
Would they be fine for this new test?

I would use an AC analysis. Produces a nice plot of frequency response that you can compare to the RIAA curve. That is what I did when I looked at the Carver C1 preamp stage.


--

David Schultz
"The cheeper the crook, the gaudier the patter." - Sam Spade

I've learned there's two anti-RIAA filters you put at the input of
preamp to be simulated: the laplace on and and RC passive one.
All my sims with RIAA premps have the passive filter at the input,
where I feed the AC signal through.
According to one of my colleagues at DIYaaudio, "Laplace is the most
accurate but can only be used in the .AC mode.
For the .Tran mode you will have to use the RC model.

As you mentioned on the use of Square Waves: you can tell quite a lot
about even an audio circuit's performance, by looking at its square
wave response. Tilt in the flat tops or ringing on the edges tells
you that something is amiss with the low frequency or high frequency
response, respectively. It's a quick-and-dirty way to judge the
broadband response.
That's what ignited my curiosity, and the recent tests with square
waves and a scope in an actual built preamp on the thread by one of
our pals.

I did try some of the SW settings that you suggested, and for now
found different responses according to the opamp being used. Now I
need someone to translate those results into actual distortions or
limits, and if possible play with the variables.

On power amps, SW testing helped trim the parallel small caps on the
feedback, which later showed as distortion changes in THD sims.

This is the passive Jung filter I use on all my RIAA sims.

(Lipshitz_and_Jung_1980)_RUS.png

Carlos

On Sun, Mar 30, 2025 at 1:21?AM Andy I via groups.io
<AI.egrps+io@...> wrote:

RIAA is intended for audio applications. Why in the world would you even think of square wave input?

Actually, you can tell quite a lot about even an audio circuit's performance, by looking at its square wave response. Tilt in the flat tops or ringing on the edges tells you that something is amiss with the low frequency or high frequency response, respectively. It's a quick-and-dirty way to judge the broadband response.

That is, of course, assuming that you include an accurate inverse RIAA network. Without that, it would be meaningless and pointless.

Andy

I never pretend or intend for someone else to do the hard work. That
is for me to try, because that way I learn.

What I did need help on that thread was with adding a HP filter for a
specific very low frequency rumble some Thorens TTs had. And I was
trying to use a two stage passive RIAA filter, as shown on Walt Jung's
book and later, and another passive/active design, also two stage,
inspired on a discrete RIAA preamp by Erno Borbely in The Audio
Amateru magazine. One pal on the thread, Marcel, helped me with that.
Designing a RIAA filter is much beyond my knowledge.

I'm not sure someone here would like to see the .asc sims, with and
without filters, as that's not a matter for this "problem solving"
site (to describe it somehow). If someone is interested, do let me
know. Going to that DIYaudio thread would be much better, and
suggestions from those that ask and respond to questions here would be
very useful there.

When I ask for help on something, everywhere, I do not expect others
to do the hard work. That is what I want to do.

Carlos

On Sun, Mar 30, 2025 at 10:04?AM Andy I via groups.io
<AI.egrps+io@...> wrote:

On Sun, Mar 30, 2025 at 08:23 AM, Carlos E. Martínez wrote:

According to one of my colleagues at DIYaaudio, "Laplace is the most
accurate but can only be used in the .AC mode.
For the .Tran mode you will have to use the RC model.

That is not quite true.

The Laplace version has the disadvantage that the Laplace transform from frequency domain to time domain sometimes does not produce the right result - but sometimes it does, and the reasons why it works or does not work are not well understood. It seems to be a problem with fine-tuning the Laplace settings in the simulator. Those settings are WINDOW, NFFT, and MTOL. This is described (tersely) in LTspice's Help. As it states there:

The time domain behavior is found from the sum of the instantaneous current(or voltage) with the convolution of the history of this current(or voltage) with the impulse response. Numerical inversion of a Laplace transfer function to the time domain impulse response is a potentially compute-bound process and a topic of current numerical research. ... This process is prone to the usual artifacts of FFT's such as spectral leakage and picket fencing that is common to discrete FFT's. ... LTspice must guess an appropriate frequency range and resolution. It is recommended that the LTspice first be allowed to make a guess at this. The length of the window and number of FFT data points used will be reported in the .log file. You can then adjust the algorithm's choices by explicitly setting nfft and window length. The reciprocal of the value of the window is the frequency resolution. The value of nfft times this resolution is the highest frequency considered. ...

In other words, it might work, but you may need to manually adjust the settings to get a good transform to the time domain. It is incorrect that "you will have to use the RC model", but it is probably better and easier if you do. An RC model with correctly calculated components has exactly the same response as the Laplace form anyway.

Your colleagues at DIYaudio are incorrect when they say that "Laplace is the most accurate." It is not more accurate than the RC version. It is just easier to calculate since you know in advance where the poles and zeros should be.


I did try some of the SW settings that you suggested, and for now
found different responses according to the opamp being used. Now I
need someone to translate those results into actual distortions or
limits, and if possible play with the variables.

Can you do that yourself? Since this is your project, you ought to be able to examine your preamps' distortions and other effects. Or, you could pay someone to do it for you. :-) Wouldn't it be better to work on it yourself? Getting assistance is good, but don't expect someone else to do all the work.

Andy